REMODE: Probabilistic, monocular dense reconstruction in real time
Open accessThe DOI matches its Crossref record; 25 of the 26 references it cites were found in a registry or an index.
Abstract
In this paper, we solve the problem of estimating dense and accurate depth maps from a single moving camera. A probabilistic depth measurement is carried out in real time on a per-pixel basis and the computed uncertainty is used to reject erroneous estimations and provide live feedback on the reconstruction progress. Our contribution is a novel approach to depth map computation that combines Bayesian estimation and recent development on convex optimization for image processing. We demonstrate that our method outperforms state-of-the-art techniques in terms of accuracy, while exhibiting high efficiency in memory usage and computing power. We call our approach REMODE (REgularized MOnocular Depth Estimation). Our CUDA-based implementation runs at 30Hz on a laptop computer and is released as open-source software.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“These representations have recently gained popularity in monocular SLAM, in conjunction with the use of direct methods [118, 184, 203], which estimate the trajectory of the robot and a 3D model directly from the intensity values of all the image pixels.”
From Past, Present, and Future of Simultaneous Localization and Mapping: Toward the Robust-Perception Age (2016) · cites this paper as [203]
Citations
Cites
Multiple View Geometry in Computer Vision
Hartley, R., & Zisserman, A. (2004) · doi:10.1017/cbo9780511811685 · Cited by 21,070
Nonlinear total variation based noise removal algorithms
Rudin, L. I., Osher, S., & Fatemi, E. (1992) · Physica D: Nonlinear Phenomena · doi:10.1016/0167-2789(92)90242-f · Cited by 15,704
A First-Order Primal-Dual Algorithm for Convex Problems with Applications to Imaging
Chambolle, A., & Pock, T. (2011) · Journal of Mathematical Imaging and Vision · doi:10.1007/s10851-010-0251-1 · Cited by 4,336
Accurate, Dense, and Robust Multiview Stereopsis
Furukawa, Y., & Ponce, J. (2010) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2009.161 · Cited by 3,019
A Comparison and Evaluation of Multi-View Stereo Reconstruction Algorithms
Seitz, S. M., Curless, B., Diebel, J., Scharstein, D., & Szeliski, R. (2006) · 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition - Volume 1 (CVPR'06) · doi:10.1109/cvpr.2006.19 · Cited by 2,457 · cited 1×
“The growing interest for dense reconstructions has renewed the attention in multi-view stereo techniques [3], [4], [5], [6], where the involved computational complexity used to prevent applications in robot perception.”
BackgroundSVO: Fast semi-direct monocular visual odometry
Forster, C., Pizzoli, M., & Scaramuzza, D. (2014) · 2014 IEEE International Conference on Robotics and Automation (ICRA) · doi:10.1109/icra.2014.6906584 · Cited by 2,205
Self-citationDTAM: Dense tracking and mapping in real-time
Newcombe, R. A., Lovegrove, S. J., & Davison, A. J. (2011) · 2011 International Conference on Computer Vision · doi:10.1109/iccv.2011.6126513 · Cited by 2,078 · cited 4×
“If, on one hand, estimating the depth independently for every pixel leads to efficient, parallel implementations, on the other hand the authors of [10], [11], [12] argued that, similar to other computer vision problems, such as image de-noising [13] and optical flow estimation [14], a smoothing step is required in ord…”
BackgroundMethodInfluentialReal-time simultaneous localisation and mapping with a single camera
Davison (2003) · Proceedings Ninth IEEE International Conference on Computer Vision · doi:10.1109/iccv.2003.1238654 · Cited by 1,733
Fast Global Minimization of the Active Contour/Snake Model
Bresson, X., Esedoḡlu, S., Vandergheynst, P., Thiran, J.-P., & Osher, S. (2007) · Journal of Mathematical Imaging and Vision · doi:10.1007/s10851-007-0002-0 · Cited by 857
Inverse Depth Parametrization for Monocular SLAM
Civera, J., Davison, A. J., & Montiel, J. (2008) · IEEE Transactions on Robotics · doi:10.1109/tro.2008.2003276 · Cited by 779
Research works printed but not linked: 1
- werlberger, 0, Motion estimation with nonlocal total variation regularization, Proc IEEE Conf on Computer Vision and Pattern Recognition 2010
Cited by
Items 201–300 of 330
Liu, C., Prior, S. D., Teacy, W. T. L., & Warner, M. B. (2016) · Advances in Mechanical Engineering · doi:10.1177/1687814016640996 · Cited by 6
“Since gE = 1⁄20, 0, 0, 1 , then it can be further simplified as equation (6)”
BackgroundModel-aided monocular visual-inertial state estimation and dense mapping
Qiu, K., & Shen, S. (2017) · IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros.2017.8205992 · Cited by 7
“REMODE [12] is another monocular solution using a similar method.”
MethodReal-Time Direct Monocular SLAM With Learning-Based Confidence Estimation
Zhang, W., Yan, Z., Xiao, G., Feng, A., & Zuo, W. (2019) · IEEE Access · doi:10.1109/access.2019.2928578 · Cited by 6
Improved ORB-SLAM Based 3D Dense Reconstruction for Monocular Endoscopic Image
Chen, W., Liao, X., Sun, Y., & Wang, Q. (2020) · International Conference on Virtual Reality and Visualization · doi:10.1109/icvrv51359.2020.00030 · Cited by 6
Review of Research on Multi-robot Visual Simultaneous Localization and Mapping
Hesheng, Y., Shuo, P., Xu, L., & Bo, H. (2022) · Journal of Mechanical Engineering · doi:10.3901/jme.2022.11.011 · Cited by 8
Handling pure camera rotation in semi-dense monocular SLAM
Zhou, Y., Yan, F., & Zhou, Z. (2017) · The Visual Computer · doi:10.1007/s00371-017-1435-0 · Cited by 13
A novel re-tracking strategy for monocular SLAM
Liu, H., Huang, W., & Wang, Z. (2017) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp.2017.7952495 · Cited by 6
“However, direct methods demand for fine initialization and high computing power (e.g., GPUs) [17, 18], which limits their applications in many fields.”
BackgroundIncremental Pose Map Optimization for Monocular Vision SLAM Based on Similarity Transformation
Liu, W., Wu, S., Zhong-bo, W., & Wu, X. (2019) · Sensors · doi:10.3390/s19224945 · Cited by 7
“In [11], a method of depth estimation based on Bayesian estimation and convex optimization image processing was proposed.”
MethodSemantic Nearest Neighbor Fields Monocular Edge Visual-OdometryPreprint
Wu, X., Benbihi, A., Richard, A., & Pradalier, C. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1904.00738 · Cited by 7
“Recent works have shown that direct methods [4] [5] [6] [7] could provide more accurate and robust motion estimation.”
Luo, X., Huang, Y. Q., Li, H., Mei, Y., Zhang, F., & Ma, M. (2025) · Computer-Aided Civil and Infrastructure Engineering · doi:10.1111/mice.70031 · Cited by 5
“Several notable technologies, such as structure from motion (SfM; Schonberger & Frahm, 2016) and regularized monocular depth estimation (Pizzoli et al., 2014), have been developed.”
COMO: Compact Mapping and Odometry
Dexheimer, E., & Davison, A. J. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-73001-6_20 · Cited by 5
Bioinspired vision-only UAV attitude rate estimation using machine learning
Merida-Floriano, M., Caballero, F., García-Morales, D., Casares, F., & Merino, L. (2017) · International Conference on Unmanned Aircraft Systems (ICUAS) · doi:10.1109/icuas.2017.7991430 · Cited by 5
“Vision-based procedures have been proposed for odometry [3], [4], [5], localization [6], [7], [8], Seville, Spain { dgarmor,fcasfer } at upo.es mapping [9] and navigation [10].”
Comparison of Power Consumption of Modern SLAM Methods on Various Datasets
Yanık, Ö. F., & Ilgın, H. A. (2021) · International Conference on Technological Advancements and Innovations (ICTAI) · doi:10.1109/ictai53825.2021.9673342 · Cited by 5
Simultaneous Localisation and Mapping With Quadric Surfaces
Laidlow, T., & Davison, A. J. (2022) · International Conference on 3D Vision (3DV) · doi:10.1109/3dv57658.2022.00037 · Cited by 6
“To address this, dense mapping systems such as DTAM [4] and REMODE [5] aim to create dense reconstructions of the scene.”
Real-time variational stereo reconstruction with applications to large-scale dense SLAM
Kuschk, G., Božič, A., & Cremers, D. (2017) · IEEE Intelligent Vehicles Symposium · doi:10.1109/ivs.2017.7995899 · Cited by 5
“Such direct Visual SLAM methods include [2], [3], [4], [5], [6], [7], [8].”
BackgroundMethodMultimodal Multispectral Imaging System for Small UAVs
Haavardsholm, T. V., Skauli, T., & Stahl, A. (2020) · IEEE Robotics and Automation Letters · doi:10.1109/lra.2020.2967301 · Cited by 5
A low-cost visual navigation and mapping system for Unmanned Aerial Vehicle using LSD-SLAM algorithm
Anandharaman, S., Sudhakaran, M., & Seyezhai, R. (2016) · Online International Conference on Green Engineering and Technologies · doi:10.1109/get.2016.7916802 · Cited by 5
Merida-Floriano, M., Caballero, F., Acedo, D., García-Morales, D., Casares, F., & Merino, L. (2019) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra.2019.8794057 · Cited by 5
“Vision-based procedures have been proposed for odometry [3], [4], [5], localization [6], [7], [8], mapping [9] and navigation [10].”
BackgroundConquering Textureless with RF-referenced Monocular Vision for MAV State Estimation
Zhang, S., Tang, S., Wang, W., Jiang, T., & Zhang, Q. (2021) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra48506.2021.9560744 · Cited by 4
“Direct methods take all the raw pixel information in images to mitigate the e ff ect of textureless scenes [24], [25].”
MethodScale-Consistent Fusion: From Heterogeneous Local Sampling to Global Immersive Rendering
Xing, W., Chen, J., Yang, Z., Wang, Q., & Guo, Y. (2022) · IEEE Transactions on Image Processing · doi:10.1109/tip.2022.3205745 · Cited by 5
“Pizzoli et al. [20] estimated the depths of pixels by searching for their correspondence points from multi-view images, and updates the point position by a robust probabilistic model.”
BackgroundReal-time Underwater 3D Reconstruction Based on Monocular Image
Hu, C., Zhu, S., & Song, W. (2022) · IEEE International Conference on Robotics and Biomimetics (ROBIO) · doi:10.1109/robio55434.2022.10011639 · Cited by 6
3D Environment Reconstruction Using Mobile Robot Platform and Monocular Vision
Dewangan, K., Saha, A., Vaiapury, K., & Dasgupta, R. (2016) · Advances in intelligent systems and computing · doi:10.1007/978-981-10-1023-1_22 · Cited by 6
Automatic Selection of Video Frames for Path Regularization and 3D Reconstruction
Pavoni, G., Dellepiane, M., Callieri, M., & Scopigno, R. (2016) · ISTI Open Portal · doi:10.2312/gch.20161376 · Cited by 4
“They are based on a concurrent segmentation of the scene [KLD ∗ 14], on the use of probabilistic models [PFS14], or on the implementation of fast Structure From Motion (SfM) [ND10].”
MethodRealtime Object-aware Monocular Depth Estimation in Onboard Systems
Lee, S., Lee, C., Kim, H., & Kim, H. J. (2021) · International Journal of Control Automation and Systems · doi:10.1007/s12555-020-0654-8 · Cited by 4
“Also, in order to deal with wide depth ranges up to infinity, we implement the filter based on inverse-depth parameterization similar to [14,15].”
BackgroundMethodA Front-End for Dense Monocular SLAM using a Learned Outlier Mask Prior
Zhang, Y., & Leonard, J. J. (2021) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra48506.2021.9561392 · Cited by 4
“In [25], this measurement uncertainty is obtained by assuming a”
BackgroundMethodLi, J., Wang, D., Li, M., Dong, T., Fu, F., & Xu, C. (2023) · IEEE Transactions on Aerospace and Electronic Systems · doi:10.1109/taes.2023.3268021 · Cited by 6
Monocular SLAM Algorithm Based on Improved Depth Map Estimation and Keyframe Selection
Kuang, H., Zhang, K., Li, R., & Liu, X. (2018) · International Conference on Measuring Technology and Mechatronics Automation (ICMTMA) · doi:10.1109/icmtma.2018.00091 · Cited by 4
“After that, a regularized method is introduced to penalize non-smooth surfaces based on the Huber norm of the gradient [14].”
MethodAdaptive Baseline Monocular Dense Mapping with Inter-Frame Depth Propagation
Wang, K., & Shen, S. (2018) · IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros.2018.8593936 · Cited by 5
“For REMODE [2], diverged estimations are masked as invalid points for a fair comparison.”
BackgroundMethodResultInfluentialSimultaneous identification of points and circles: structure from motion system in industry scenes
Ni, T., Shi, Y., Sun, A., & Ju, B. (2020) · Pattern Analysis and Applications · doi:10.1007/s10044-020-00889-2 · Cited by 4
“The variance 2 i is calculated by single pixel photometric disparity in [24].”
MethodSimultaneous localization and mapping using monocular direct method
Meng, C., Guo, B., & Liu, X. (2016) · International Conference on Control, Automation, Robotics and Vision (ICARCV) · doi:10.1109/icarcv.2016.7838785 · Cited by 4
“Second, a novel probabilistic approach in [10]to incorporate noise on the estimated depth maps.”
Yeh, C.-F., Chao-yang, T., Chen, T.-C., Chang, M., Hsieh, C.-C., Liu, R.-S., Tang, K., & Lo, C. (2024) · doi:10.36227/techrxiv.171216655.57408835/v1 · Cited by 4
Augmented Reality Enabled Telepresence Framework for Intuitive Robotic Teleoperation
Xu, Z., Lin, Z., Liu, H., Ren, Q., & Gao, A. (2023) · IEEE International Conference on Robotics and Biomimetics (ROBIO) · doi:10.1109/robio58561.2023.10354631 · Cited by 3
“The output information, containing frame, pose, and depth estimation, is input into the second dense reconstruction, which uses the regularized monocular depth estimation (REMODE) [26] to obtain the dense point cloud.”
BackgroundMethodInfluentialMVS-GS: High-Quality 3D Gaussian Splatting Mapping via Online Multi-View Stereo
Lee, B., Park, J., Giang, K. T., Jo, S., & Song, S. (2025) · IEEE Access · doi:10.1109/access.2025.3583156 · Cited by 7
“Dense simultaneous localization and mapping (SLAM) [10] [11] [12] tackles the challenge of online 3D modeling.”
A direct visual-inertial sensor fusion approach in multi-state constraint Kalman filter
Gui, J., & Gu, D. (2015) · Chinese Control Conference (CCC) · doi:10.1109/chicc.2015.7260595 · Cited by 3
Liu, C., Nash, J., & Prior, S. D. (2015) · ePrints Soton (University of Southampton) · doi:10.5281/zenodo.1109889 · Cited by 3
Computer Vision for Micro Air Vehicles
Brockers, R., Humenberger, M., Kuwata, Y., Matthies, L. H., & Weiß, S. (2014) · Advances in computer vision and pattern recognition · doi:10.1007/978-3-319-09387-1_4 · Cited by 3
An Improved ORB-SLAM2 With Refined Depth Estimation
Tang, K., Yuan, J., Sun, Q., Zhang, X., & Gao, H. (2019) · IEEE International Conference on Real-time Computing and Robotics (RCAR) · doi:10.1109/rcar47638.2019.9044034 · Cited by 3
“Forster and Pizzoli [8] [9] proposed the concept of depth filter and deduced the depth filter based on the Uniform-Gaussian mixture distribution.”
BackgroundLiDAR and Camera Calibration Using Pyramid and Checkerboard Calibrators
Zhang, B., Zheng, Y., Zhang, Z., & He, Q. (2023) · IEEE International Conference on Big Data Analytics (ICBDA) · doi:10.1109/icbda57405.2023.10104660 · Cited by 3
“In the area of camera reconstruction, [2] and [3] employed a progressive reconstruction technique to continuously fuse 3D information using images from the subsequent moment while utilizing the Kalman filter concept for depth reconstruction.”
MethodLiu, C., Prior, S. D., & Scanlan, J. P. (2016) · Unmanned Systems · doi:10.1142/s2301385016500059 · Cited by 3
“where matrix B and C are defined in (12) and (13) respectively, and T = (T ? 1 , T ? 2 , T ? 3 , T ? 4 ) > is desired thrust vector, which represents the desired thrust command to each rotor.”
BackgroundAutonomous Multirrotor Design and Simulation for Search and Rescue Missions
Alarcon, C., & Jamett, M. (2018) · IEEE International Conference on Automatica · doi:10.1109/ica-acca.2018.8609813 · Cited by 3
“M. Jamett is with the Electrical Engineering Department of Universidad de Santiago de Chile, Chile (e-mail: marcela.jamett@usach.cl). desarrolla la técnica de odometría visual semi-directa (Semidirect Visual Odometry, SVO) [2], la cual permite obtener la posición y orientación del MAV sin depender de un sistema de loc…”
BackgroundLightweight Low-Cost UAV Radar Terrain Mapping
Ye, E., Shaker, G. S. A., & Melek, W. (2019) · European Conference on Antennas and Propagation · Cited by 3
Towards the Probabilistic Fusion of Learned Priors into Standard Pipelines for 3D Reconstruction
Laidlow, T., Czarnowski, J., Nicastro, A., Clark, R., & Leutenegger, S. (2020) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra40945.2020.9197001 · Cited by 4
Structure-Aware SLAM with Planes in Man-Made Environment
Zhang, J., Zeng, G., & Zha, H. (2017) · Communications in computer and information science · doi:10.1007/978-981-10-7305-2_41 · Cited by 3
Mixed Probability Inverse Depth Estimation Based on Probabilistic Graph Model
Liu, W., Wu, S., Wu, X., & Li, K. (2019) · IEEE Access · doi:10.1109/access.2019.2920278 · Cited by 3
Edge-Aware Monocular Dense Depth Estimation with Morphology
Li, Z., Zhu, X., Yu, H., Zhang, Q., & Jiang, Y. (2021) · Proceedings - International Conference on Pattern Recognition/Proceedings/International Conference on Pattern Recognition · doi:10.1109/icpr48806.2021.9412578 · Cited by 5
“Most Virtual SLAM methods produce sparse depth maps, however, some SLAM methods [9], [10] compute all pixels in the image and attempt to reconstruct dense depth maps.”
MethodPragma-Oriented Parallelization of the Direct Sparse Odometry SLAM Algorithm
Pereira, C., Falcão, G., & Alexandre, L. A. (2019) · Euromicro International Conference on Parallel, Distributed and Network-Based Processing (PDP) · doi:10.1109/empdp.2019.8671561 · Cited by 3
IMRL: An Improved Inertial-Aided KLT Feature Tracker
Quan, M., Mu, B., & Chai, Z. (2019) · IEEE Conference on Cybernetics and Intelligent Systems · doi:10.1109/cis-ram47153.2019.9095829 · Cited by 3
“The calculation of depth measurement uncertainty is motivated by the method of [14].”
MethodMDEAN: Multi-View Disparity Estimation with an Asymmetric Network
Pei, Z., Wen, D., Zhang, Y., Ma, M., Guo, M., Zhang, X., & Yang, Y. (2020) · Electronics · doi:10.3390/electronics9060924 · Cited by 3
“[28] propose a method combining bayesian estimation and convex optimization, which can accurately predict in real time the scene depth of each pixel of an image sequence.”
BackgroundMethodSözer, O., & Kumbasar, T. (2023) · Drones · doi:10.3390/drones7040278 · Cited by 2
Online 3D Scene Reconstruction Using Neural Object Priors
Chabal, T., Chen, S., Ponce, J., & Schmid, C. (2025) · International Conference on 3D Vision (3DV) · doi:10.1109/3dv66043.2025.00072 · Cited by 4
“Other scene representations include point clouds [57] and surfels [66, 87] which are fast to compute but remain discrete, a limitation for a number of downstream applications.”
Exploiting Structure-from-Motion for Robust Vision-Based Map Matching for Aircraft Surface Movement
Choate, D., & Rife, J. H. (2025) · Proceedings of the Satellite Division's International Technical Meeting (Online)/Proceedings of the Satellite Division's International Technical Meeting (CD-ROM) · doi:10.33012/2025.20382 · Cited by 2
EyeGlove for Inspection and Measurement in Confined Environments
Sun, E., Axinte, D. A., Alatorre, D., & Dong, X. (2022) · IEEE Transactions on Industrial Electronics · doi:10.1109/tie.2022.3167144 · Cited by 2
“The epi-polar constraint can be solved by the triangulation method [23] (2) Refer to (2), the corresponding matched pixel P R of a pixel P L in the Cam L frame, must lie in the epi-polar line of the Cam R frame [24], [25].”
MethodImam, J. M., Mokhtar, N. B., Hatta, S. F. W. M., Rashdan, M., Salman, M., Khandaker, M. U., & Islam, M. A. (2024) · IEEE Sensors Journal · doi:10.1109/jsen.2024.3433471 · Cited by 3
“The major limitation of DLV is, that if the water is too deep, the acoustic signal does not reach the bottom and the DVL cannot measure SOG [42].”
BackgroundInfluentialPassive Initialization Method Based on Motion Characteristics for Monocular SLAM
Yu, Y., Xiong, J., She, X., Liu, C., Yang, C., & Li, J. (2019) · Complexity · doi:10.1155/2019/8176489 · Cited by 2
“SVO’s initialization is similar to that of PTAM, except that SVO integrates an additional assumption that the motion direction is perpendicular to the photographed plane, as SVO is originally designed for rotor UAV use.”
MethodWhat We Can Learn From the Primate’s Visual System
Krüger, N., Zillich, M., Janssen, P., & Buch, A. G. (2015) · Künstliche Intell. · doi:10.1007/s13218-014-0345-9 · Cited by 2
“and optic flow [57,73,84,18,61], is still a challenging research topic.”
BackgroundMetric online monocular SLAM by using a known reference
Li, Y., Wang, S., Yang, D., & Sun, D. (2016) · World Congress on Intelligent Control and Automation · doi:10.1109/wcica.2016.7578496 · Cited by 2
Dense and Occlusion-Robust Multi-view Stereo for Unstructured Videos
Wei, J., Resch, B., & Lensch, H. P. A. (2016) · Canadian Conference on Computer and Robot Vision · doi:10.1109/crv.2016.42 · Cited by 3
“Other methods [8], [9] have been introduced for fast and dense recovery of 3D scenes from video streams.”
MethodProbabilistic Depth Map Model for Rotation-Only Camera Motion in Semi-Dense Monocular SLAM
Zhou, Y., Yan, F., & Zhou, Z. (2016) · International Conference on Virtual Reality and Visualization · doi:10.1109/icvrv.2016.11 · Cited by 2
“Furthermore, as proposed in [10] and [11], the posterior in (2) can be approximated by the product of a Gaussian distribution for the depth and a Beta distribution for the probability of good measurement: 2 2 ( , |a ,b , , ) ( | , ) ( | , ) k k k k k k k k q d Beta a b N d π μ σ π μ σ ∧ ∧ = (3)”
BackgroundReal-time enhancement of sparse 3D maps using a parallel segmentation scheme based on superpixels
Cruz-Martínez, C., Martínez-Carranza, J., & Mayol-Cuevas, W. W. (2017) · Journal of Real-Time Image Processing · doi:10.1007/s11554-017-0707-2 · Cited by 2
Time-efficient dense visual 12-DoF state estimator using RGB-D camera
Kim, C., Lee, S.-I., Kim, P., & Kim, H. J. (2017) · International Conference on Ubiquitous Robots and Ambient Intelligence · doi:10.1109/urai.2017.7992902 · Cited by 2
“Additionally, it also takes away several sophisticated issues on VO, such as scale drift and depth propagation of general RGB cameras [12].”
BackgroundMap Point Optimization in Keyframe-Based SLAM using Covisibility Graph and Information Fusion
Concha, E. K. T., Pittol, D., Westhauser, R. S., Kolberg, M., Maffei, R. D. Q., & Prestes, E. (2019) · International Conference on Advanced Robotics (ICAR) · doi:10.1109/icar46387.2019.8981653 · Cited by 2
“a covisibility graph [12] and an algorithm of information fusion [14] to maintain a refined map in keyframe-based monocular SLAM process.”
BackgroundMethodLearning Depth from Monocular Sequence with Convolutional LSTM Network
Yeh, C., Huang, Y.-P., Lin, C., & Lin, M.-H. (2019) · Advances in intelligent systems and computing · doi:10.1007/978-3-030-29029-0_48 · Cited by 2
Towards Collaborative Obstacle Avoidance using Small UAS in Indoor Environments
Duran, D., Johnson, M., & Stansbury, R. S. (2020) · IEEE/ION Position Location and Navigation Symposium · doi:10.1109/plans46316.2020.9110141 · Cited by 2
A method of Monocular Visual Odometry Combining Feature points and Pixel Gradient for Dynamic Scene
Li, P., Li, H., & Chen, B. (2020) · IEEE Vehicular Technology Conference · doi:10.1109/vtc2020-spring48590.2020.9129019 · Cited by 2
“The SLAM algorithm combining the dotted line feature method and the LK(Lucas-Kanade) [1] optical flow, although the optical flow method still also relies on the strong assumption that the brightness of the light is constant [15], and it is based on the visual odometer system under small amplitude motion [10].”
MethodMonocular adaptive inverse depth filtering algorithm based on Gaussian model
Xu, C., Wu, C., Qu, D., Sun, H., Song, J., & Wang, X. (2020) · Chinese Control and Decision Conference (CCDC) · doi:10.1109/ccdc49329.2020.9164819 · Cited by 2
Autonomous vision-based landing of UAV’s on unstructured terrains
Chatzikalymnios, E., & Μουστάκας, Κ. (2021) · IEEE International Conference on Autonomous Systems (ICAS) · doi:10.1109/icas49788.2021.9551180 · Cited by 2
Multi-scale Direct Sparse Visual Odometry for Large-Scale Natural Environment
Wu, X., & Pradalier, C. (2018) · International Conference on 3D Vision (3DV) · doi:10.1109/3dv.2018.00021 · Cited by 2
“For direct dense mapping, [17] [13] [16], the photometric error with spacial geometry prior formulations are developed to accurately reconstruct the dense depth map and the camera motion.”
MethodReal-time monocular 3D perception with ORB-Features
Ji, B., & Cao, Q. (2018) · Industrial Robot the international journal of robotics research and application · doi:10.1108/ir-06-2018-0115 · Cited by 2
Virtual testbed for monocular visual navigation of small unmanned aircraft systems
Kim, K. C., Leishman, R. C., & Nykl, S. L. (2020) · The Journal of Defense Modeling and Simulation Applications Methodology Technology · doi:10.1177/1548512920954545 · Cited by 3
“Follow on methods 17 fused SVO outputs with IMU measurements to autonomously control a quadrotor and execute a trajectory through an onboard processor while providing a real-time dense 3D map of the traversed area using REMODE on an offboard computer.”
MethodA Monocular SLAM System with Mask Loop Closing
Han, B., & Xu, L. (2020) · Chinese Control and Decision Conference (CCDC) · doi:10.1109/ccdc49329.2020.9164164 · Cited by 2
Dense 3D-Reconstruction from Monocular Image Sequences for Computationally Constrained UAS
Domnik, M., Proença, P. F., Delaune, J. H., Thiem, J., & Brockers, R. (2021) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv48630.2021.00186 · Cited by 3
“An example is REMODE (REgularized MOnocular Depth Estimation) [25] which is a real-time temporal fusion approach based on Bayesian estimation that heavily uses GPU parallelization in order to reach acceptable execution times.”
BackgroundMethodCombining LiDAR Scan Matching with Stereo Visual Odometry Using Curvefusion
Du, S., Li, X., Lauterbach, H., Borrmann, D., & Nüchter, A. (2021) · International Conference on Computer, Control and Robotics (ICCCR) · doi:10.1109/icccr49711.2021.9349385 · Cited by 2
“Direct methods operate directly on pixel intensities without calculating feature points [8].”
MethodDense monocular Simultaneous Localization and Mapping by direct surfel optimization
Trabes, E., Avila, L., Dondo, J., & Páez, C. S. (2021) · Journal of Applied Research and Technology · doi:10.22201/icat.24486736e.2021.19.6.991 · Cited by 1
“REMODE (Pizzoli et al., 2014) uses an epipolar depth search together with a probabilistic measurement merging, again with a regularization at the end of the process.”
MethodInfluentialDeep-DSO: Improving Mapping of Direct Sparse Odometry Using CNN-Based Single-Image Depth Estimation
Herrera-Granda, E. P., Torres, J. C., Herrera-Granda, I. D., Lucio-Naranjo, J., Rosales, A., Revelo-Fuelagán, J., & Peluffo-Ordóñez, D. H. (2025) · Mathematics · doi:10.3390/math13203330 · Cited by 4
Robust direct visual inertial odometry via entropy-based relative pose estimation
Gui, J., Gu, D., & Hu, H. (2015) · IEEE International Conference on Mechatronics and Automation (ICMA) · doi:10.1109/icma.2015.7237603 · Cited by 1
“have been proposed: In [12], [2], fully dense depth maps are built, depending on the computational ability of state-of-art GPU, while it is very hard to launch on an easily practical usage.”
BackgroundDensification of Semi-Dense Reconstructions for Novel View Generation of Live Scenes
Baričević, D., Höllerer, T., & Turk, M. (2017) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv.2017.99 · Cited by 1
“[25], using the authors’ implementation [1].”
Monocular Dense Reconstruction Based on Direct Sparse Odometry
Mao, L., Wu, J., Zhang, J., & Chen, S. (2017) · Communications in computer and information science · doi:10.1007/978-981-10-7305-2_38 · Cited by 1
The Software Tool for Comparison and Configuration of Nonlinear Optimization Techniques in Orb-Slam
Boris, M., & Afanasyev, I. M. (2019) · International Conference on Developments in eSystems Engineering (DeSE) · doi:10.1109/dese.2019.00095 · Cited by 1
“Examples are DTAM [12], REMODE [13] and LSD-SLAM [14].”
MethodTriangulation Precision Prediction for Road Elements Updating in High-Definition Map
Pei, X., Sun, P., Yang, M., Yang, D., & Sui, G. (2021) · International Conference on Control, Automation and Robotics (ICCAR) · doi:10.1109/iccar52225.2021.9463500 · Cited by 1
“Matia Pizzoli introduces REMODE concept [8], which gives a detailed analysis about the triangulation precision and develop depth filter for real-time 3D reconstruction.”
Background3D Crane Site Reconstruction with Lifting Path Displaying by a Top-view Camera
Wang, Y., Xi, N., Bi, S., & Zhang, Q. (2022) · Creative Construction e-Conference · doi:10.3311/ccc2022-001 · Cited by 1
MVSOP: A new framework for integrating MVS into SLAM based on ORB and PF
Zhang, Z., Tang, F., Li, Y., & Feng, X. (2023) · Youth Academic Annual Conference of Chinese Association of Automation (YAC) · doi:10.1109/yac59482.2023.10401431 · Cited by 1
“…(cid:1) our (cid:1) method (cid:1) can (cid:1) integrate (cid:1) the (cid:1) dense (cid:1) reconstruction (cid:1) with (cid:1) traditional (cid:1) visual (cid:1) SLAM (cid:1) frameworks, (cid:1) the (cid:1) algorithm (cid:1) [19] (cid:1) and (cid:1) our (cid:1) method (cid:1) are (cid:1) compared.”
MethodInfluentialTES-CVIDS: A Transmission Efficient Sub-Map Based Collaborative Dense VI-SLAM Framework
Zhang, T., Zhang, L., Zhang, F., Zhao, S., & Zhou, Y. (2024) · IEEE Transactions on Intelligent Vehicles · doi:10.1109/tiv.2024.3385452 · Cited by 1
“In [36], the first fully distributed multi-robot system for dense metric-semantic SLAM, namely Kimera-Multi, was presented.”
Real-time Monocular Dense Mapping of Small Scenes with ORB Features
Ji, B., Cao, Q., & Zhu, X. (2018) · International Conference on Advanced Robotics and Mechatronics (ICARM) · doi:10.1109/icarm.2018.8610747 · Cited by 1
Baraian, A., & Nedevschi, S. (2019) · IEEE International Conference on Intelligent Computer Communication and Processing (ICCP) · doi:10.1109/iccp48234.2019.8959696 · Cited by 2
“Opposed to that are dense methods that take into consideration almost all pixels from an image [12], but their performance significantly drops.”
BackgroundPersian sign language detection based on normalized depth image information
Rajabi, S., & Mousavinia, A. (2019) · Journal of Control · doi:10.29252/joc.13.3.29 · Cited by 1
NBVC: A Benchmark for Depth Estimation from Narrow-Baseline Video Clips
Mordohai, P., Batsos, K., Makadia, A., & Snavely, N. (2020) · IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros45743.2020.9340817 · Cited by 1
“While impressive results in terms of processing speed and visual quality have been achieved [15], [16], [17], [18], [19], [20], [21], [22], [23], evaluation is almost exclusively qualitative.”
BackgroundZhong, Z., & Mehltretter, M. (2021) · ISPRS annals of the photogrammetry, remote sensing and spatial information sciences · doi:10.5194/isprs-annals-v-2-2021-17-2021 · Cited by 1
“According to Vogiatzis and Hernandez (2011) and Pizzoli et al. (2014), a depth sensor produces two types of measurement: (1) a good measurement that is unimodally distributed around the correct depth or (2) an outlier measurement that is drawn from a Uniform distribution defined on a certain…”
BackgroundHigh-quality dense SLAM approach from deep learning and monocular-inertial measurements
Duan, Z., Hu, B., Yu, H., & Chen, C. (2021) · International Conference on Algorithms, High Performance Computing and Artificial Intelligence (AHPCAI) · doi:10.1117/12.2626462 · Cited by 1
A Novel Fusion Perception Method of Unmanned Driving Based on Distributed Roadside Information
Zhang, Q., Qian, L., Liu, L., Ding, L., & Yang, F. (2022) · International Conference on Advances in Computer Technology, Information Science and Communications (CTISC) · doi:10.1109/ctisc54888.2022.9849823 · Cited by 1
PiMPeR: Piecewise Dense 3D Reconstruction from Multi-View and Multi-Illumination ImagesPreprint
Sabzevari, R., Murino, V., & Bue, A. D. (2015) · arXiv (Cornell University) · doi:10.48550/arxiv.1503.04598 · Cited by 1
“Recently, Pizzoli et al. [4] and Engel et al. [5] showed that only high gradient textures could be used as reliable observations of the scene for 3D reconstruction, and that, to recover 3D position of the scene parts with homogeneous textures, strong prior information such as surface smoothness is…”
BackgroundVision-Based Tracking, Odometry and Control for UAV Autonomy
Fu, C. (2015) · UPM Digital Archive (Technical University of Madrid) · doi:10.20868/upm.thesis.38382 · Cited by 1
Robust Semi-Direct Monocular Visual Odometry Using Edge and Illumination-Robust CostPreprint
Wu, X., & Pradalier, C. (2019) · arXiv (Cornell University) · Cited by 1
Positioning and Mapping for Aerial Robots using on-board Perception for Autonomous Missions
Bavle, H. (2019) · UPM Digital Archive (Technical University of Madrid) · doi:10.20868/upm.thesis.57710 · Cited by 4
Pose estimation using visual entropy
Gui, J., Gu, D., & Hu, H. (2015) · International Conference on Automation and Computing (ICAC) · doi:10.1109/iconac.2015.7313960
“Methods in [5], [6] built a fully dense depth maps on per pixel basis and the computational ability of state-of-art GPU were adopted.”
BackgroundJoint Recursive Monocular Filtering of Camera Motion and Disparity Map
Berger, J., & Schnörr, C. (2016) · Lecture notes in computer science · doi:10.1007/978-3-319-45886-1_19
“Monocular methods [9, 1, 13, 12, 20, 19, 5] benefit from less calibration effort in comparison to stereo methods, but suffer from a peculiarity of the mathematical setup that prevents to reconstruct the scale of the scene uniquely.”
BackgroundAn accurate and efficient monocular mixed match SLAM
Dai, W.-C., Zhang, Y., & Li, P. (2017) · Cybersecurity and Cyberforensics Conference · doi:10.23919/chicc.2017.8028440
Monocular Visual-Inertial SLAM and Self Calibration for Long Term Autonomy
Keivan, N. (2017) · CU Scholar (University of Colorado Boulder)
A method of antenna barrier detection and depth estimation
Zhu, B., Jia, Q., Ye, P., Han, S., & Sun, H. (2017) · International Conference on Systems and Informatics (ICSAI) · doi:10.1109/icsai.2017.8248280
“We refer to [12], which consider the depth of the pixel as a normal distribution, and the inlier or not as BETA distribution.”
BackgroundMethodInfluentialGCP-SLAM: LSD-SLAM with Learning-Based Confidence Estimation
Feng, A., Zhang, W., Yan, Z., & Zuo, W. (2018) · Lecture notes in computer science · doi:10.1007/978-3-319-75786-5_22
Monocular dense reconstruction by depth estimation fusion
Chen, T., Ding, W., Zhang, D., & Liu, X. (2018) · Chinese Control and Decision Conference (CCDC) · doi:10.1109/ccdc.2018.8407902
“Dense reconstruction methods [4, 5, 1, 2] reconstruct surfaces and require GPU acceleration due to the enormous computational cost.”
Method
Metadata
- Authors
- Pizzoli, M., Forster, C., & Scaramuzza, D.
- Affiliations
- University of Zurich (CH)
- ORCID
- Scaramuzza, Davide
- Venue
- 2014 IEEE International Conference on Robotics and Automation (ICRA)
- Publisher
- IEEE
- Published
- 2014-05
- Pages
- 2609-2616
- ISSN
- 1050-4729 (ISSN-L)
- OpenAlex
- W2093659073
- Cited by
- 330
- References
- 26
- Citations by year
- 2026: 3 · 2025: 11 · 2024: 10 · 2023: 16 · 2022: 26 · 2021: 33 · 2020: 36 · 2019: 48 · 2018: 39 · 2017: 47 · 2016: 33 · 2015: 20 · 2014: 8
- FWCI
- 19.36
- Type
- conference paper
- Language
- en
- Open access
- green
- www.zora.uzh.ch
- Licence
- other-oa
- Locations
- Proceedings - IEEE International Conference on Robotics and Automation/Proceedings (publishedVersion)
- Zurich Open Repository and Archive (University of Zurich) (acceptedVersion, other-oa)
- Infoscience (Ecole Polytechnique Fédérale de Lausanne) (submittedVersion)
- Infoscience (Ecole Polytechnique Fédérale de Lausanne) (submittedVersion, cc-by-nc-nd)
- Zurich Open Repository and Archive (University of Zurich) (submittedVersion, other-oa)
- Zurich Open Repository and Archive (University of Zurich)
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Vision and Imaging
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Robotics and Sensor-Based Localization
- Physical Sciences › Engineering › Media Technology › Image Processing Techniques and Applications
- Keywords
- monocular depth estimation, dense reconstruction, uncertainty estimation, Bayesian estimation, convex optimization, real-time depth estimation, CUDA
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.